• radio_button_unchecked

Importing required libraries¶

Converting date columns value in date time datatype¶

Data Reading¶

Sno Date Time State/UnionTerritory ConfirmedIndianNational ConfirmedForeignNational Cured Deaths Confirmed
0 1 2020-01-30 6:00 PM Kerala 1 0 0 0 1
1 2 2020-01-31 6:00 PM Kerala 1 0 0 0 1
2 3 2020-02-01 6:00 PM Kerala 2 0 0 0 2
3 4 2020-02-02 6:00 PM Kerala 3 0 0 0 3
4 5 2020-02-03 6:00 PM Kerala 3 0 0 0 3

Converting date column to datetime datatype

Renaming columns

Index(['Sno', 'Date', 'Time', 'States', 'ConfirmedIndianNational',
       'ConfirmedForeignNational', 'Cured', 'Deaths', 'Confirmed'],
      dtype='object')

Checking null values

Sno                         0
Date                        0
Time                        0
States                      0
ConfirmedIndianNational     0
ConfirmedForeignNational    0
Cured                       0
Deaths                      0
Confirmed                   0
dtype: int64
(18110, 9)

Dropping unwanted columns

Date Time States Cured Deaths Confirmed
0 2020-01-30 6:00 PM Kerala 0 0 1
1 2020-01-31 6:00 PM Kerala 0 0 1
2 2020-02-01 6:00 PM Kerala 0 0 2
3 2020-02-02 6:00 PM Kerala 0 0 3
4 2020-02-03 6:00 PM Kerala 0 0 3

Calculating active value

Date Time States Cured Deaths Confirmed Active
18105 2021-08-11 8:00 AM Telangana 638410 3831 650353 8112
18106 2021-08-11 8:00 AM Tripura 77811 773 80660 2076
18107 2021-08-11 8:00 AM Uttarakhand 334650 7368 342462 444
18108 2021-08-11 8:00 AM Uttar Pradesh 1685492 22775 1708812 545
18109 2021-08-11 8:00 AM West Bengal 1506532 18252 1534999 10215
Date Time States Cured Deaths Confirmed Active
0 2020-01-30 6:00 PM Kerala 0 0 1 1
1 2020-01-31 6:00 PM Kerala 0 0 1 1
2 2020-02-01 6:00 PM Kerala 0 0 2 2
3 2020-02-02 6:00 PM Kerala 0 0 3 3
4 2020-02-03 6:00 PM Kerala 0 0 3 3

Dataset Information

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 18110 entries, 0 to 18109
Data columns (total 7 columns):
 #   Column     Non-Null Count  Dtype         
---  ------     --------------  -----         
 0   Date       18110 non-null  datetime64[ns]
 1   Time       18110 non-null  object        
 2   States     18110 non-null  object        
 3   Cured      18110 non-null  int64         
 4   Deaths     18110 non-null  int64         
 5   Confirmed  18110 non-null  int64         
 6   Active     18110 non-null  int64         
dtypes: datetime64[ns](1), int64(4), object(2)
memory usage: 990.5+ KB

After manipulating covid_summary data there are 2 categorical, 1 datetime, 1 float and 4 int values

Selecting maximum date data

Timestamp('2021-08-11 00:00:00')
Cured Deaths Confirmed Active
States
Andaman and Nicobar Islands 7412 129 7548 7
Andhra Pradesh 1952736 13564 1985182 18882
Arunachal Pradesh 47821 248 50605 2536
Assam 559684 5420 576149 11045
Bihar 715352 9646 725279 281

Looking at 2021 total cured, deaths, confirmed and active data

Cured Deaths Confirmed Active
0 31220981 429179 32036511 386351

Storing confirmed, death and cured data<.b>

Date Confirmed Deaths Cured
0 2020-01-30 1 0 0
1 2020-01-31 1 0 0
2 2020-02-01 2 0 0
3 2020-02-02 3 0 0
4 2020-02-03 3 0 0

Cured, Deaths, Confirmed Trend bar

Mar 2020May 2020Jul 2020Sep 2020Nov 2020Jan 2021Mar 2021May 2021Jul 2021010M20M30M40M50M60M
ConfirmedDeathsRecovered
plotly-logomark

Method to plot bar chart

Top 10 Death States

020k40k60k80k100k120k140kChhattisgarhAndhra PradeshPunjabKeralaWest BengalUttar PradeshDelhiTamil NaduKarnatakaMaharashtra
Top 10 States with most deathsNumber of deaths(In Thousands)State Name
plotly-logomark

Top 10 Confirmed States

01M2M3M4M5M6MOdishaChhattisgarhDelhiWest BengalUttar PradeshAndhra PradeshTamil NaduKarnatakaKeralaMaharashtra
Top 10 Indian States (Confirmed Cases)Number of Confirmed cases (in Thousands)States Name
plotly-logomark

Top 10 Cured States

01M2M3M4M5M6MOdishaChhattisgarhDelhiWest BengalUttar PradeshAndhra PradeshTamil NaduKarnatakaKeralaMaharashtra
Top 10 States (Cured Cases)Number of Cured cases (in Thousands)States Name
plotly-logomark

Since we know that maharashtra is most affected state of India lets look at other data of Maharashtra. To proceed forward lets read vaccination data

Observing Vaccination data on Maharashtra State

Updated On State Total Doses Administered Sessions Sites First Dose Administered Second Dose Administered Male (Doses Administered) Female (Doses Administered) Transgender (Doses Administered) ... 18-44 Years (Doses Administered) 45-60 Years (Doses Administered) 60+ Years (Doses Administered) 18-44 Years(Individuals Vaccinated) 45-60 Years(Individuals Vaccinated) 60+ Years(Individuals Vaccinated) Male(Individuals Vaccinated) Female(Individuals Vaccinated) Transgender(Individuals Vaccinated) Total Individuals Vaccinated
7840 11/08/2021 West Bengal NaN NaN NaN NaN NaN NaN NaN NaN ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
7841 12/08/2021 West Bengal NaN NaN NaN NaN NaN NaN NaN NaN ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
7842 13/08/2021 West Bengal NaN NaN NaN NaN NaN NaN NaN NaN ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
7843 14/08/2021 West Bengal NaN NaN NaN NaN NaN NaN NaN NaN ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN
7844 15/08/2021 West Bengal NaN NaN NaN NaN NaN NaN NaN NaN ... NaN NaN NaN NaN NaN NaN NaN NaN NaN NaN

5 rows × 24 columns

Index(['Date', 'State', 'Total Doses Administered', 'Sessions', ' Sites ',
       'First Dose Administered', 'Second Dose Administered',
       'Male (Doses Administered)', 'Female (Doses Administered)',
       'Transgender (Doses Administered)', ' Covaxin (Doses Administered)',
       'CoviShield (Doses Administered)', 'Sputnik V (Doses Administered)',
       'AEFI', '18-44 Years (Doses Administered)',
       '45-60 Years (Doses Administered)', '60+ Years (Doses Administered)',
       '18-44 Years(Individuals Vaccinated)',
       '45-60 Years(Individuals Vaccinated)',
       '60+ Years(Individuals Vaccinated)', 'Male(Individuals Vaccinated)',
       'Female(Individuals Vaccinated)', 'Transgender(Individuals Vaccinated)',
       'Total Individuals Vaccinated', 'Total Vaccinatons'],
      dtype='object')
Date                                     0
State                                    0
Total Doses Administered                 6
Sessions                                 6
 Sites                                   6
First Dose Administered                  6
Second Dose Administered                 6
Male (Doses Administered)                6
Female (Doses Administered)              6
Transgender (Doses Administered)         6
 Covaxin (Doses Administered)            6
CoviShield (Doses Administered)          6
Sputnik V (Doses Administered)         131
AEFI                                    65
18-44 Years (Doses Administered)       166
45-60 Years (Doses Administered)       166
60+ Years (Doses Administered)         166
18-44 Years(Individuals Vaccinated)    112
45-60 Years(Individuals Vaccinated)    111
60+ Years(Individuals Vaccinated)      111
Male(Individuals Vaccinated)           212
Female(Individuals Vaccinated)         212
Transgender(Individuals Vaccinated)    212
Total Individuals Vaccinated            52
Total Vaccinatons                        6
dtype: int64
Date                                   0
State                                  0
Total Doses Administered               0
Sessions                               0
 Sites                                 0
First Dose Administered                0
Second Dose Administered               0
Male (Doses Administered)              0
Female (Doses Administered)            0
Transgender (Doses Administered)       0
 Covaxin (Doses Administered)          0
CoviShield (Doses Administered)        0
Sputnik V (Doses Administered)         0
AEFI                                   0
18-44 Years (Doses Administered)       0
45-60 Years (Doses Administered)       0
60+ Years (Doses Administered)         0
18-44 Years(Individuals Vaccinated)    0
45-60 Years(Individuals Vaccinated)    0
60+ Years(Individuals Vaccinated)      0
Male(Individuals Vaccinated)           0
Female(Individuals Vaccinated)         0
Transgender(Individuals Vaccinated)    0
Total Individuals Vaccinated           0
Total Vaccinatons                      0
dtype: int64

Total vaccination provided on Maharashtra

16/01/202119/01/202122/01/202125/01/202128/01/202131/01/202103/02/202106/02/202109/02/202112/02/202115/02/202118/02/202121/02/202124/02/202127/02/202102/03/202105/03/202108/03/202111/03/202114/03/202117/03/202120/03/202123/03/202126/03/202129/03/202101/04/202104/04/202107/04/202110/04/202113/04/202116/04/202119/04/202122/04/202125/04/202128/04/202101/05/202104/05/202107/05/202110/05/202113/05/202116/05/202119/05/202122/05/202125/05/202128/05/202131/05/202103/06/202106/06/202109/06/202112/06/202115/06/202118/06/202121/06/202124/06/202127/06/202130/06/202103/07/202106/07/202109/07/202112/07/202115/07/202118/07/202121/07/202124/07/202127/07/202130/07/202102/08/202105/08/202108/08/202111/08/202114/08/2021010M20M30M40M
Vaccination till date in MaharashtraTotal VaccinatonsDate
plotly-logomark

Total CoviShield vaccination provided on Maharashtra

16/01/202119/01/202122/01/202125/01/202128/01/202131/01/202103/02/202106/02/202109/02/202112/02/202115/02/202118/02/202121/02/202124/02/202127/02/202102/03/202105/03/202108/03/202111/03/202114/03/202117/03/202120/03/202123/03/202126/03/202129/03/202101/04/202104/04/202107/04/202110/04/202113/04/202116/04/202119/04/202122/04/202125/04/202128/04/202101/05/202104/05/202107/05/202110/05/202113/05/202116/05/202119/05/202122/05/202125/05/202128/05/202131/05/202103/06/202106/06/202109/06/202112/06/202115/06/202118/06/202121/06/202124/06/202127/06/202130/06/202103/07/202106/07/202109/07/202112/07/202115/07/202118/07/202121/07/202124/07/202127/07/202130/07/202102/08/202105/08/202108/08/202111/08/202114/08/2021010M20M30M40M
CoviShield Administered in MahrashtraCoviShield (Doses Administered)Date
plotly-logomark

Total Covaxin vaccination provided on Maharashtra

16/01/202119/01/202122/01/202125/01/202128/01/202131/01/202103/02/202106/02/202109/02/202112/02/202115/02/202118/02/202121/02/202124/02/202127/02/202102/03/202105/03/202108/03/202111/03/202114/03/202117/03/202120/03/202123/03/202126/03/202129/03/202101/04/202104/04/202107/04/202110/04/202113/04/202116/04/202119/04/202122/04/202125/04/202128/04/202101/05/202104/05/202107/05/202110/05/202113/05/202116/05/202119/05/202122/05/202125/05/202128/05/202131/05/202103/06/202106/06/202109/06/202112/06/202115/06/202118/06/202121/06/202124/06/202127/06/202130/06/202103/07/202106/07/202109/07/202112/07/202115/07/202118/07/202121/07/202124/07/202127/07/202130/07/202102/08/202105/08/202108/08/202111/08/202114/08/202102M4M6M
Covaxin Administered in MaharashtraCovaxin (Doses Administered)Date
plotly-logomark

We used the Prophet library, which was created by Facebook and is used for Time Series Forecasting, to make our predictions. Prophet is a time series data forecasting procedure based on an additive model that fits non-linear trends with yearly, weekly, and daily seasonality, as well as holiday effects. It works best with time series with strong seasonal effects and historical data from multiple seasons.

FBProphet¶

Making model to predict and compare the confirmed data¶
Date Confirmed
555 2021-08-07 31895385
556 2021-08-08 31934455
557 2021-08-09 31969954
558 2021-08-10 31998158
559 2021-08-11 32036511
<fbprophet.forecaster.Prophet at 0x1b939e37eb0>

Future 30 days date prediction

ds
585 2021-09-06
586 2021-09-07
587 2021-09-08
588 2021-09-09
589 2021-09-10

Forcasting India confirmation data

ds trend yhat_lower yhat_upper trend_lower trend_upper additive_terms additive_terms_lower additive_terms_upper weekly weekly_lower weekly_upper multiplicative_terms multiplicative_terms_lower multiplicative_terms_upper yhat
0 2020-01-30 -1.945281e+05 -1.836273e+06 1.420716e+06 -1.945281e+05 -1.945281e+05 20670.809283 20670.809283 20670.809283 20670.809283 20670.809283 20670.809283 0.0 0.0 0.0 -1.738573e+05
1 2020-01-31 -1.912244e+05 -1.769051e+06 1.338023e+06 -1.912244e+05 -1.912244e+05 15736.475838 15736.475838 15736.475838 15736.475838 15736.475838 15736.475838 0.0 0.0 0.0 -1.754879e+05
2 2020-02-01 -1.879207e+05 -1.765816e+06 1.367934e+06 -1.879207e+05 -1.879207e+05 10539.935153 10539.935153 10539.935153 10539.935153 10539.935153 10539.935153 0.0 0.0 0.0 -1.773807e+05
3 2020-02-02 -1.846169e+05 -1.748840e+06 1.402066e+06 -1.846169e+05 -1.846169e+05 4514.591868 4514.591868 4514.591868 4514.591868 4514.591868 4514.591868 0.0 0.0 0.0 -1.801023e+05
4 2020-02-03 -1.813132e+05 -1.682445e+06 1.387868e+06 -1.813132e+05 -1.813132e+05 -4087.407399 -4087.407399 -4087.407399 -4087.407399 -4087.407399 -4087.407399 0.0 0.0 0.0 -1.854006e+05
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
585 2021-09-06 4.053997e+07 3.880602e+07 4.206546e+07 4.032867e+07 4.077778e+07 -4087.407399 -4087.407399 -4087.407399 -4087.407399 -4087.407399 -4087.407399 0.0 0.0 0.0 4.053588e+07
586 2021-09-07 4.070569e+07 3.912970e+07 4.244741e+07 4.047697e+07 4.095049e+07 -18976.383761 -18976.383761 -18976.383761 -18976.383761 -18976.383761 -18976.383761 0.0 0.0 0.0 4.068672e+07
587 2021-09-08 4.087142e+07 3.914108e+07 4.246668e+07 4.062752e+07 4.112666e+07 -28398.020981 -28398.020981 -28398.020981 -28398.020981 -28398.020981 -28398.020981 0.0 0.0 0.0 4.084302e+07
588 2021-09-09 4.103715e+07 3.928715e+07 4.261594e+07 4.078187e+07 4.130496e+07 20670.809283 20670.809283 20670.809283 20670.809283 20670.809283 20670.809283 0.0 0.0 0.0 4.105782e+07
589 2021-09-10 4.120287e+07 3.964374e+07 4.299325e+07 4.092711e+07 4.148270e+07 15736.475837 15736.475837 15736.475837 15736.475837 15736.475837 15736.475837 0.0 0.0 0.0 4.121861e+07

590 rows × 16 columns

The below image shows the basic prediction. The light blue is the uncertainty level(yhat_upper and yhat_lower), the dark blue is the prediction(yhat) and the black line is the original data.

Basic line plot Forecast India Confirmation

Jan 2020Apr 2020Jul 2020Oct 2020Jan 2021Apr 2021Jul 2021010M20M30M40M
1w1m6m1yallyds
plotly-logomark

The images below depict the time series data's trends and seasonality (within a year).

Forecasting India confirmation weekly and trend data

In rapid trend growths, changepoints are added to indicate the time. The dotted red lines indicate when there was a significant change in the passenger trend.

Forecasting India confirmed data with changepoints

Predicting the number of cases till Oct 2021, there will be more than 45 millions people who will be affected by coronavirus

Making model to predict and compare the cured data¶
Date Cured
555 2021-08-07 31055861
556 2021-08-08 31099771
557 2021-08-09 31139457
558 2021-08-10 31180968
559 2021-08-11 31220981
ds y
555 2021-08-07 31055861
556 2021-08-08 31099771
557 2021-08-09 31139457
558 2021-08-10 31180968
559 2021-08-11 31220981
ds
585 2021-09-06
586 2021-09-07
587 2021-09-08
588 2021-09-09
589 2021-09-10
ds trend yhat_lower yhat_upper trend_lower trend_upper additive_terms additive_terms_lower additive_terms_upper weekly weekly_lower weekly_upper multiplicative_terms multiplicative_terms_lower multiplicative_terms_upper yhat
0 2020-01-30 -1.158339e+05 -1.468481e+06 1.305507e+06 -1.158339e+05 -1.158339e+05 20451.412190 20451.412190 20451.412190 20451.412190 20451.412190 20451.412190 0.0 0.0 0.0 -9.538248e+04
1 2020-01-31 -1.139421e+05 -1.452425e+06 1.271701e+06 -1.139421e+05 -1.139421e+05 13607.673366 13607.673366 13607.673366 13607.673366 13607.673366 13607.673366 0.0 0.0 0.0 -1.003344e+05
2 2020-02-01 -1.120502e+05 -1.471254e+06 1.166625e+06 -1.120502e+05 -1.120502e+05 7765.453372 7765.453372 7765.453372 7765.453372 7765.453372 7765.453372 0.0 0.0 0.0 -1.042848e+05
3 2020-02-02 -1.101584e+05 -1.501450e+06 1.329662e+06 -1.101584e+05 -1.101584e+05 2085.288507 2085.288507 2085.288507 2085.288507 2085.288507 2085.288507 0.0 0.0 0.0 -1.080731e+05
4 2020-02-03 -1.082666e+05 -1.595541e+06 1.254691e+06 -1.082666e+05 -1.082666e+05 -8067.857442 -8067.857442 -8067.857442 -8067.857442 -8067.857442 -8067.857442 0.0 0.0 0.0 -1.163345e+05
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
585 2021-09-06 3.953173e+07 3.816574e+07 4.088229e+07 3.927164e+07 3.974831e+07 -8067.857442 -8067.857442 -8067.857442 -8067.857442 -8067.857442 -8067.857442 0.0 0.0 0.0 3.952366e+07
586 2021-09-07 3.970386e+07 3.818226e+07 4.098763e+07 3.943088e+07 3.993058e+07 -13767.146379 -13767.146379 -13767.146379 -13767.146379 -13767.146379 -13767.146379 0.0 0.0 0.0 3.969010e+07
587 2021-09-08 3.987600e+07 3.850566e+07 4.125089e+07 3.957728e+07 4.011847e+07 -22074.823614 -22074.823614 -22074.823614 -22074.823614 -22074.823614 -22074.823614 0.0 0.0 0.0 3.985393e+07
588 2021-09-09 4.004814e+07 3.863782e+07 4.140736e+07 3.973356e+07 4.030339e+07 20451.412190 20451.412190 20451.412190 20451.412190 20451.412190 20451.412190 0.0 0.0 0.0 4.006859e+07
589 2021-09-10 4.022027e+07 3.892233e+07 4.163199e+07 3.989079e+07 4.049674e+07 13607.673365 13607.673365 13607.673365 13607.673365 13607.673365 13607.673365 0.0 0.0 0.0 4.023388e+07

590 rows × 16 columns

The below image shows the basic prediction. The light blue is the uncertainty level(yhat_upper and yhat_lower), the dark blue is the prediction(yhat) and the black line is the original data.

Basic line plot Forecast India Cured

Jan 2020Apr 2020Jul 2020Oct 2020Jan 2021Apr 2021Jul 2021010M20M30M40M
1w1m6m1yallyds
plotly-logomark

The images below depict the time series data's trends and seasonality (within a year).

Forecasting India Cured weekly and trend data

In rapid trend growths, changepoints are added to indicate the time. The dotted red lines indicate when there was a significant change in the passenger trend.

Forecasting India cured data with changepoints

Predicting number of cured till Oct 2021 will be more than 40 million

Making model to predict and compare the cured data¶
Date Deaths
555 2021-08-07 427371
556 2021-08-08 427862
557 2021-08-09 428309
558 2021-08-10 428682
559 2021-08-11 429179
ds
585 2021-09-06
586 2021-09-07
587 2021-09-08
588 2021-09-09
589 2021-09-10
ds trend yhat_lower yhat_upper trend_lower trend_upper additive_terms additive_terms_lower additive_terms_upper weekly weekly_lower weekly_upper multiplicative_terms multiplicative_terms_lower multiplicative_terms_upper yhat
0 2020-01-30 -2199.302798 -15937.214557 12346.712912 -2199.302798 -2199.302798 274.316839 274.316839 274.316839 274.316839 274.316839 274.316839 0.0 0.0 0.0 -1924.985959
1 2020-01-31 -2149.567673 -16837.756422 11142.038380 -2149.567673 -2149.567673 194.220341 194.220341 194.220341 194.220341 194.220341 194.220341 0.0 0.0 0.0 -1955.347332
2 2020-02-01 -2099.832549 -17360.268628 12467.742619 -2099.832549 -2099.832549 144.504564 144.504564 144.504564 144.504564 144.504564 144.504564 0.0 0.0 0.0 -1955.327985
3 2020-02-02 -2050.097425 -16740.243007 11859.622955 -2050.097425 -2050.097425 37.969937 37.969937 37.969937 37.969937 37.969937 37.969937 0.0 0.0 0.0 -2012.127488
4 2020-02-03 -2000.362300 -15966.728200 12515.753614 -2000.362300 -2000.362300 -93.461547 -93.461547 -93.461547 -93.461547 -93.461547 -93.461547 0.0 0.0 0.0 -2093.823847
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
585 2021-09-06 537128.772850 522271.298147 552985.488646 534022.192114 540276.061650 -93.461547 -93.461547 -93.461547 -93.461547 -93.461547 -93.461547 0.0 0.0 0.0 537035.311303
586 2021-09-07 539499.684080 523855.939795 555215.448787 536149.029965 542938.547049 -260.003341 -260.003341 -260.003341 -260.003341 -260.003341 -260.003341 0.0 0.0 0.0 539239.680738
587 2021-09-08 541870.595309 526704.353432 556166.780907 538315.742208 545528.792062 -297.546792 -297.546792 -297.546792 -297.546792 -297.546792 -297.546792 0.0 0.0 0.0 541573.048517
588 2021-09-09 544241.506538 528487.239743 559420.830376 540492.722305 548073.173042 274.316839 274.316839 274.316839 274.316839 274.316839 274.316839 0.0 0.0 0.0 544515.823377
589 2021-09-10 546612.417768 531514.250211 562217.849584 542630.376465 550703.356801 194.220341 194.220341 194.220341 194.220341 194.220341 194.220341 0.0 0.0 0.0 546806.638109

590 rows × 16 columns

The below image shows the basic prediction. The light blue is the uncertainty level(yhat_upper and yhat_lower), the dark blue is the prediction(yhat) and the black line is the original data.

Basic line plot Forecast India Deaths

Jan 2020Apr 2020Jul 2020Oct 2020Jan 2021Apr 2021Jul 20210100k200k300k400k500k
1w1m6m1yallyds
plotly-logomark

The images below depict the time series data's trends and seasonality (within a year).

Forecasting India Deaths weekly and trend data

In rapid trend growths, changepoints are added to indicate the time. The dotted red lines indicate when there was a significant change in the passenger trend.

Forecasting India deaths data with changepoints

Predicting number of deaths till Oct 2021 will be more than 600k. We can see that current situation of the India is not under control.

Accuracy testing by taking kerela states¶

Copying origingal dataset

ds Time States y Deaths Confirmed Active
0 2020-01-30 6:00 PM Kerala 0 0 1 1
1 2020-01-31 6:00 PM Kerala 0 0 1 1
2 2020-02-01 6:00 PM Kerala 0 0 2 2
3 2020-02-02 6:00 PM Kerala 0 0 3 3
4 2020-02-03 6:00 PM Kerala 0 0 3 3

Removing unwanted state data

array(['Kerala', 'Telengana', 'Delhi', 'Rajasthan', 'Uttar Pradesh',
       'Haryana', 'Ladakh', 'Tamil Nadu', 'Karnataka', 'Maharashtra',
       'Punjab', 'Jammu and Kashmir', 'Andhra Pradesh', 'Uttarakhand',
       'Odisha', 'Puducherry', 'West Bengal', 'Chhattisgarh',
       'Chandigarh', 'Gujarat', 'Himachal Pradesh', 'Madhya Pradesh',
       'Bihar', 'Manipur', 'Mizoram', 'Andaman and Nicobar Islands',
       'Goa', 'Unassigned', 'Assam', 'Jharkhand', 'Arunachal Pradesh',
       'Tripura', 'Nagaland', 'Meghalaya',
       'Dadra and Nagar Haveli and Daman and Diu',
       'Cases being reassigned to states', 'Sikkim', 'Daman & Diu',
       'Lakshadweep', 'Telangana', 'Dadra and Nagar Haveli',
       'Himanchal Pradesh', 'Karanataka'], dtype=object)

Observing maximum and minimum data

Timestamp('2021-08-11 00:00:00')
Timestamp('2020-01-30 00:00:00')

Seperating training and testing data

(12566, 7)
(5540, 7)
ds Time States y Deaths Confirmed Active
0 2020-01-30 6:00 PM Kerala 0 0 1 1
1 2020-01-31 6:00 PM Kerala 0 0 1 1
2 2020-02-01 6:00 PM Kerala 0 0 2 2
3 2020-02-02 6:00 PM Kerala 0 0 3 3
4 2020-02-03 6:00 PM Kerala 0 0 3 3

Forecasting Kerala data with prophet prediction

ds trend yhat_lower yhat_upper trend_lower trend_upper additive_terms additive_terms_lower additive_terms_upper weekly weekly_lower weekly_upper multiplicative_terms multiplicative_terms_lower multiplicative_terms_upper yhat
0 2020-01-30 -9.703408e+01 -3.234766e+03 3.590867e+03 -9.703408e+01 -9.703408e+01 112.166231 112.166231 112.166231 112.166231 112.166231 112.166231 0.0 0.0 0.0 1.513216e+01
1 2020-01-31 -9.393595e+01 -3.244146e+03 3.412608e+03 -9.393595e+01 -9.393595e+01 94.673872 94.673872 94.673872 94.673872 94.673872 94.673872 0.0 0.0 0.0 7.379257e-01
2 2020-02-01 -9.083782e+01 -3.570888e+03 3.385396e+03 -9.083782e+01 -9.083782e+01 37.147985 37.147985 37.147985 37.147985 37.147985 37.147985 0.0 0.0 0.0 -5.368983e+01
3 2020-02-02 -8.773969e+01 -3.445288e+03 3.357780e+03 -8.773969e+01 -8.773969e+01 56.673879 56.673879 56.673879 56.673879 56.673879 56.673879 0.0 0.0 0.0 -3.106581e+01
4 2020-02-03 -8.464156e+01 -3.308185e+03 3.446105e+03 -8.464156e+01 -8.464156e+01 13.688051 13.688051 13.688051 13.688051 13.688051 13.688051 0.0 0.0 0.0 -7.095351e+01
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
431 2021-04-05 1.185537e+06 1.173031e+06 1.198146e+06 1.174239e+06 1.198111e+06 13.688051 13.688051 13.688051 13.688051 13.688051 13.688051 0.0 0.0 0.0 1.185550e+06
432 2021-04-06 1.190764e+06 1.177212e+06 1.204733e+06 1.178478e+06 1.203946e+06 -150.791339 -150.791339 -150.791339 -150.791339 -150.791339 -150.791339 0.0 0.0 0.0 1.190614e+06
433 2021-04-07 1.195992e+06 1.181800e+06 1.210919e+06 1.182797e+06 1.209764e+06 -163.558678 -163.558678 -163.558678 -163.558678 -163.558678 -163.558678 0.0 0.0 0.0 1.195829e+06
434 2021-04-08 1.201220e+06 1.186905e+06 1.217140e+06 1.187258e+06 1.215659e+06 112.166231 112.166231 112.166231 112.166231 112.166231 112.166231 0.0 0.0 0.0 1.201332e+06
435 2021-04-09 1.206447e+06 1.190917e+06 1.222322e+06 1.191428e+06 1.221715e+06 94.673872 94.673872 94.673872 94.673872 94.673872 94.673872 0.0 0.0 0.0 1.206542e+06

436 rows × 16 columns

Basic forcasted line plot of Kerala

Accuracy testing

(436, 2)
(154, 2)
32402424 1517167.9528269023
0.9531773316457157

Here we got 95 percentage of accuracy on forcasted data which is good

Checking forecasting accuracy for top 5 populated states¶

['Uttar Pradesh', 'Maharashtra', 'Bihar', 'West Bengal', 'Madhya Pradesh']
Uttar Pradesh: 0.9810909368947742
Maharashtra: 0.9159660536396657
Bihar: 0.9770128540772085
West Bengal: 0.9860752640040191
Madhya Pradesh: 0.9830944209538308
Here we can see the forecasting accuracy percentage of top 5 populated areas are more than 90 percentage.¶